Rising oil prices, concerns about climate change, and future energy supplies have contributed to growing interest in the use of liquid biofuels in the transport sector which, in turn, has driven large-scale land acquisitions in developing countries for biofuel feedstock production, mainly jatropha. The increasing trend of land acquisition for biofuels has led to the widespread debate about food versus biofuel because of the perceived competition for land and water. To avoid the food versus fuel debate, the use of “marginal” land for biofuel feedstock production (jatropha) has emerged as a dominant narrative. But both the availability and suitability of “marginal” land for commercial level jatropha production is not well understood/examined, especially in Africa. Using a case study of large-scale jatropha plantation in Ethiopia, this paper examines the process of land identification for jatropha investments, and the agronomic performance of large-scale jatropha plantation on so-called marginal land. Although it has been argued that jatropha can be grown well on marginal land without irrigation, and thus does not compete for land and water or displace food production from agricultural land, this study indicates that moisture stress is the key factor in the failure of many large-scale jatropha plantations in Ethiopia.